5x CEO/CIO/CISO. Top 5 globally in AI & cybersecurity. Forward-deployed with C-suites & boards. Military veteran, author, keynote speaker. Signal over noise.
The UK just wrote "loss of oversight or control of an AI agent" into its cyber bill. It doesn't define it yet. Your regulator will ask one question first: show me what the agent did. If you can't, you already lost control.
Authority should follow proven judgment, not a good demo.
So what's one decision you would never hand to an AI agent today, and what would it have to prove first?
https://t.co/5l2hMVU3Ih
The hardest call I make isn't what to decide. It's who I trust to decide when I'm not in the room.
Every company giving AI agents authority today is making that same call. Most don't realize it.
Agentic AI becomes dangerous when authority outruns control.
If you give an agent a senior operator's access and an intern's oversight, you've built the failure in on day one.
@SoonerSteve@1ForeverUSA There is no defending Mateer. He is absolutely horrible and he continually proves it over and over. Arbuckle is just as horrible.